nvme-mpath: delete disk after last connection
PATCH v2 1/1 nvme-mpath: delete disk after last connection
From: Keith Busch <[email protected]>
The multipath code currently deletes the disk only after all references to it are dropped rather than when the last path to that disk is lost. This has been reported to cause problems with some usage, like MD RAID.
Delete the disk when the last path is gone. This is the same behavior we currently have with non-multipathed nvme devices.
The following is just a simple example that demonstrates what is currently observed using a simple nvme loop back (loop setup file not shown):
# nvmetcli restore loop.json [ 31.156452] nvmet: adding nsid 1 to subsystem testnqn1 [ 31.159140] nvmet: adding nsid 1 to subsystem testnqn2
# nvme connect -t loop -n testnqn1 -q hostnqn [ 36.866302] nvmet: creating controller 1 for subsystem testnqn1 for NQN hostnqn. [ 36.872926] nvme nvme3: new ctrl: "testnqn1"
# nvme connect -t loop -n testnqn1 -q hostnqn [ 38.227186] nvmet: creating controller 2 for subsystem testnqn1 for NQN hostnqn. [ 38.234450] nvme nvme4: new ctrl: "testnqn1"
# nvme connect -t loop -n testnqn2 -q hostnqn [ 43.902761] nvmet: creating controller 3 for subsystem testnqn2 for NQN hostnqn. [ 43.907401] nvme nvme5: new ctrl: "testnqn2"
# nvme connect -t loop -n testnqn2 -q hostnqn [ 44.627689] nvmet: creating controller 4 for subsystem testnqn2 for NQN hostnqn. [ 44.641773] nvme nvme6: new ctrl: "testnqn2"
# mdadm --create /dev/md0 --level=mirror --raid-devices=2 /dev/nvme3n1 /dev/nvme5n1 [ 53.497038] md/raid1:md0: active with 2 out of 2 mirrors [ 53.501717] md0: detected capacity change from 0 to 66060288
# cat /proc/mdstat
Personalities : [raid1]
md0 : active raid1 nvme5n1[1] nvme3n1[0]
64512 blocks super 1.2 [2/2] [UU]Now delete all paths to one of the namespaces:
# echo 1 > /sys/class/nvme/nvme3/delete_controller # echo 1 > /sys/class/nvme/nvme4/delete_controller
We have no path, but mdstat says:
# cat /proc/mdstat
Personalities : [raid1]
md0 : active (auto-read-only) raid1 nvme5n1[1]
64512 blocks super 1.2 [2/1] [_U]And this is reported to cause a problem.
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.
And mdstat shows only the viable members:
# cat /proc/mdstat
Personalities : [raid1]
md0 : active (auto-read-only) raid1 nvme5n1[1]
64512 blocks super 1.2 [2/1] [_U]--- drivers/nvme/host/core.c | 2 +- drivers/nvme/host/multipath.c | 1 - drivers/nvme/host/nvme.h | 11 ++++++++++- 3 files changed, 11 insertions(+), 3 deletions(-)
drivers/nvme/host/core.c +1 -1
@@ -542,7 +542,7 @@ static void nvme_free_ns_head(struct kref *ref)struct nvme_ns_head *head = container_of(ref, struct nvme_ns_head, ref); -nvme_mpath_remove_disk (head); +nvme_mpath_put_disk (head);ida_simple_remove (&head->subsys->ns_ida, head->instance);cleanup_srcu_struct (&head->srcu);nvme_put_subsystem (head->subsys);
drivers/nvme/host/multipath.c +0 -1
@@ -729,7 +729,6 @@ void nvme_mpath_remove_disk(struct nvme_ns_head *head)*/ head->disk->queue =NULL ; } -put_disk (head->disk); }int nvme_mpath_init (struct nvme_ctrl *ctrl,struct nvme_id_ctrl *id)
drivers/nvme/host/nvme.h +10 -1
@@ -693,12 +693,18 @@ bool nvme_mpath_clear_current_path(struct nvme_ns *ns);void nvme_mpath_clear_ctrl_paths (struct nvme_ctrl *ctrl);struct nvme_ns *nvme_find_path (struct nvme_ns_head *head); +static inline void nvme_mpath_put_disk (struct nvme_ns_head *head) +{ +if (head->disk) +put_disk (head->disk); +} +static inline void nvme_mpath_check_last_path (struct nvme_ns *ns) {struct nvme_ns_head *head = ns->head;if (head->disk &&list_empty (&head->list)) -kblockd_schedule_work (&head->requeue_work); +nvme_mpath_remove_disk (head); }static inline void nvme_trace_bio_complete (struct request *req)
@@ -738,6 +744,9 @@ static inline void nvme_mpath_add_disk(struct nvme_ns *ns, struct nvme_id_ns *id) { } +static inline void nvme_mpath_put_disk (struct nvme_ns_head *head) +{ +}static inline void nvme_mpath_remove_disk (struct nvme_ns_head *head) { }
Signature · 7 lines
2.26.2 _______________________________________________ Linux-nvme mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-nvme
Christoph Hellwig
maintainerQuoted 4 lines · Hannes Reinecke
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.
So how is this going to work for e.g. a case where the device disappear due to resets or fabrics connection problems? This now directly teards down the device.
_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
A message that is not in the archive (sent off-list or not yet received): [email protected]
Christoph Hellwig
maintainerQuoted 1 line · Hannes Reinecke
<div dir='auto'><div dir="auto"><div dir="auto"><br></div><div><br><div class="elided-text">Am 20.04.2021 10:05 schrieb Christoph Hellwig <[email protected]>:<br type="attribution"><blockquote style="margin:0 0 0 0.8ex;border-left:1px #ccc solid;padding-left:1ex"><p dir="ltr">On Fri, Apr 16, 2021 at 08:24:11AM +0200, Hannes Reinecke wrote:
I'm unable to read this garbage unfortunately.
_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Quoted 10 lines · Christoph Hellwig
Quoted 4 lines · Hannes Reinecke
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.So how is this going to work for e.g. a case where the device disappear due to resets or fabrics connection problems? This now directly teards down the device.
Yes, that is correct; the nshead will be removed once the last path is _removed_. But key point here is that once the system finds itself in that situation it's impossible to recover, as the refcounts are messed. Even a manual connect call with the same parameter will _not_ restore operation, but rather result in a new namespace.
So with this patch we change from stalled I/O (where the user has to
reboot the machine to restore operation) to I/O errors (giving the user at least a _chance_ to recover).
I can easily consider adding a 'queue_if_no_path' option to allow I/O to be held even once all paths are disconnected, but that will be enother patch.
Cheers,
Hannes
Signature · 9 lines
Dr. Hannes Reinecke Kernel Storage Architect [email protected] +49 911 74053 688 SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer _______________________________________________ Linux-nvme mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-nvme
Keith Busch
maintainerQuoted 7 lines · Hannes Reinecke
Quoted 6 lines · Christoph Hellwig
Quoted 4 lines · Hannes Reinecke
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.
Quoted 6 lines
Quoted 4 lines
So how is this going to work for e.g. a case where the device disappear due to resets or fabrics connection problems? This now directly teards down the device.
Yes, that is correct; the nshead will be removed once the last path is _removed_.
The end result is also how non-multipath nvme behaves, so I think that's what users have come to expect.
Quoted 4 lines
But key point here is that once the system finds itself in that situation it's impossible to recover, as the refcounts are messed. Even a manual connect call with the same parameter will _not_ restore operation, but rather result in a new namespace.
I haven't looked at this yet, but is it really not possible to restore the original namespace upon the reestablished connection?
Quoted 6 lines
So with this patch we change from stalled I/O (where the user has to reboot the machine to restore operation) to I/O errors (giving the user at least a _chance_ to recover).
I can easily consider adding a '
queue_if_no_path' option to allow I/O to be held even once all paths are disconnected, but that will be enother patch.
_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Christoph Hellwig
maintainerQuoted 12 lines · Keith Busch
Quoted 11 lines · Hannes Reinecke
Quoted 10 lines · Christoph Hellwig
Quoted 4 lines · Hannes Reinecke
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.So how is this going to work for e.g. a case where the device disappear due to resets or fabrics connection problems? This now directly teards down the device.
Quoted 5 lines
Quoted 2 lines
Yes, that is correct; the nshead will be removed once the last path is _removed_.
The end result is also how non-multipath nvme behaves, so I think that's what users have come to expect.
I'm not sure that is what users expect. At least the SCSI multipath setups I've worked do not expect it and ensure the queue_if_no_path option is set.
Quoted 7 lines
Quoted 4 lines
But key point here is that once the system finds itself in that situation it's impossible to recover, as the refcounts are messed. Even a manual connect call with the same parameter will _not_ restore operation, but rather result in a new namespace.
I haven't looked at this yet, but is it really not possible to restore the original namespace upon the reestablished connection?
It is possible, and in fact is what we do.
_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Quoted 16 lines · Christoph Hellwig
Quoted 15 lines · Keith Busch
Quoted 13 lines · Hannes Reinecke
Quoted 10 lines · Christoph Hellwig
Quoted 4 lines · Hannes Reinecke
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.So how is this going to work for e.g. a case where the device disappear due to resets or fabrics connection problems? This now directly teards down the device.
Yes, that is correct; the nshead will be removed once the last path is _removed_.
Quoted 7 lines
Quoted 2 lines
The end result is also how non-multipath nvme behaves, so I think that's what users have come to expect.
I'm not sure that is what users expect. At least the SCSI multipath setups I've worked do not expect it and ensure the
queue_if_no_pathoption is set.
Yes, sure. And as I said, I'm happy to implement this option for NVMe, too. But that is _not_ what this patch is about.
NVMe since day one has _removed_ the namespace if the controller goes away (ie if you do a PCI hotplug). So customer rightly expects this behaviour to continue.
And this is what the patch does; _aligning_ the behaviour from multipathed to non-multipathed controllers when the last path is gone.
Non-multipathed (ie CMIC==0) controllers will remove the namespace once the last _reference_ to that namespace drops (ie the PCI hotplug case). Multipathed (ie CMIC!=0) controllers will remove the namespace once the last _opener_ goes away.
The refcount is long gone by that time.
Quoted 10 lines
Quoted 7 lines
Quoted 4 lines
But key point here is that once the system finds itself in that situation it's impossible to recover, as the refcounts are messed. Even a manual connect call with the same parameter will _not_ restore operation, but rather result in a new namespace.
I haven't looked at this yet, but is it really not possible to restore the original namespace upon the reestablished connection?
It is possible, and in fact is what we do.
It is _not_ once the namespace is mounted.
Or MD has claimed the device.
And the problem is that the refcount already _is_ zero, so we are already in teardown. We're just waiting for the reference to the gendisk to drop. Which is never will, as we would have to umount (or detach) the device for that, but I/O is still pending which cannot be flushed, so that'll fail. And if we try to connect the same namespace again, nvme_find_ns_head() will not return the existing ns_head as the refcount is zero. Causing a new ns_head to be created.
If you manage to get this working with the current code please show me from the testcase in the description what we should have done differently.
Cheers,
Hannes
Signature · 9 lines
Dr. Hannes Reinecke Kernel Storage Architect [email protected] +49 911 74053 688 SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer _______________________________________________ Linux-nvme mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-nvme
Keith Busch
maintainerQuoted 60 lines · Hannes Reinecke
Quoted 16 lines · Christoph Hellwig
Quoted 15 lines · Keith Busch
Quoted 13 lines · Hannes Reinecke
Quoted 10 lines · Christoph Hellwig
Quoted 4 lines · Hannes Reinecke
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.So how is this going to work for e.g. a case where the device disappear due to resets or fabrics connection problems? This now directly teards down the device.
Yes, that is correct; the nshead will be removed once the last path is _removed_.
Quoted 7 lines
Quoted 2 lines
The end result is also how non-multipath nvme behaves, so I think that's what users have come to expect.
I'm not sure that is what users expect. At least the SCSI multipath setups I've worked do not expect it and ensure the
queue_if_no_pathoption is set.Yes, sure. And as I said, I'm happy to implement this option for NVMe, too. But that is _not_ what this patch is about.
NVMe since day one has _removed_ the namespace if the controller goes away (ie if you do a PCI hotplug). So customer rightly expects this behaviour to continue.
And this is what the patch does; _aligning_ the behaviour from multipathed to non-multipathed controllers when the last path is gone.
Non-multipathed (ie CMIC==0) controllers will remove the namespace once the last _reference_ to that namespace drops (ie the PCI hotplug case). Multipathed (ie CMIC!=0) controllers will remove the namespace once the last _opener_ goes away.
The refcount is long gone by that time.Quoted 10 lines
Quoted 7 lines
Quoted 4 lines
But key point here is that once the system finds itself in that situation it's impossible to recover, as the refcounts are messed. Even a manual connect call with the same parameter will _not_ restore operation, but rather result in a new namespace.
I haven't looked at this yet, but is it really not possible to restore the original namespace upon the reestablished connection?
It is possible, and in fact is what we do.
It is _not_ once the namespace is mounted.
Or MD has claimed the device.
And the problem is that the refcount already _is_ zero, so we are already in teardown. We're just waiting for the reference to the gendisk to drop. Which is never will, as we would have to umount (or detach) the device for that, but I/O is still pending which cannot be flushed, so that'll fail. And if we try to connect the same namespace again,nvme_find_ns_head()will not return the existingns_headas the refcount is zero.
Causing a newns_headto be created.
Quoted 2 lines
If you manage to get this working with the current code please show me from the testcase in the description what we should have done differently.
I see what you mean. I think we can detangle this, but it's not as straight forward as I hoped.
As far as *this* patch goes, I think you and I are aligned on the behavior, and I still think it's good.
_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Quoted 9 lines · Christoph Hellwig
Quoted 4 lines · Keith Busch
The end result is also how non-multipath nvme behaves, so I think that's
what users have come to expect.I'm not sure that is what users expect. At least the SCSI multipath setups I've worked do not expect it and ensure the
queue_if_no_pathoption is set.
For dm-multipath, this only holds for the case that all paths are failed. But if the last path is actually *deleted*, multipathd flushes the map (tries to, at least).
Martin
_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Quoted 12 lines · Keith Busch
Quoted 11 lines · Hannes Reinecke
Quoted 10 lines · Christoph Hellwig
Quoted 4 lines · Hannes Reinecke
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.So how is this going to work for e.g. a case where the device disappear due to resets or fabrics connection problems? This now directly teards down the device.
Quoted 6 lines
Quoted 2 lines
Yes, that is correct; the nshead will be removed once the last path is _removed_.
The end result is also how non-multipath nvme behaves, so I think that's what users have come to expect.
Nope. It doesn't.
This patch precisely _aligns_ the behaviour between multipathed (or, to be precise, namespaces with CMIC != 0) and non-multipathed ones.
When CMIC == 0 this patch doesn't change the behaviour at all.
Quoted 8 lines
Quoted 4 lines
But key point here is that once the system finds itself in that situation it's impossible to recover, as the refcounts are messed. Even a manual connect call with the same parameter will _not_ restore operation, but rather result in a new namespace.
I haven't looked at this yet, but is it really not possible to restore the original namespace upon the reestablished connection?
No. As mentioned, refcount is messed up.
With the current behaviour, once the last path drops, the namespace will _vanish_ from the list of namespaces. Once you reconnect the original namespace 'magically' reappears in 'nvme list', but with size '0', _and_ the new namespace beside it:
Needless to say, the current opener (mount, MD, you name it) still refer to the original namespace, and you can't correct the situation as I/O is still pending you have no way to abort it.
(Funny, though; it was you who suggested the patch in the first place; feels kinda strange to explain the patch to its author ...)
Cheers,
Hannes
Signature · 9 lines
Dr. Hannes Reinecke Kernel Storage Architect [email protected] +49 911 74053 688 SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer _______________________________________________ Linux-nvme mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-nvme
Keith Busch
maintainerQuoted 19 lines · Hannes Reinecke
Now delete all paths to one of the namespaces:
# echo 1 > /sys/class/nvme/nvme3/delete_controller # echo 1 > /sys/class/nvme/nvme4/delete_controllerWe have no path, but mdstat says:
# cat /proc/mdstat Personalities : [raid1] md0 : active (auto-read-only) raid1 nvme5n1[1] 64512 blocks super 1.2 [2/1] [_U]And this is reported to cause a problem.
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.
Quoted 6 lines
And mdstat shows only the viable members:
# cat /proc/mdstat Personalities : [raid1] md0 : active (auto-read-only) raid1 nvme5n1[1] 64512 blocks super 1.2 [2/1] [_U]
I think there's a copy-paste error in this change log (probably my mistake). The first mdstat should have shown nvme3n1 was still there, right?
_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Quoted 22 lines · Keith Busch
Quoted 21 lines · Hannes Reinecke
Now delete all paths to one of the namespaces:
# echo 1 > /sys/class/nvme/nvme3/delete_controller # echo 1 > /sys/class/nvme/nvme4/delete_controllerWe have no path, but mdstat says:
# cat /proc/mdstat Personalities : [raid1] md0 : active (auto-read-only) raid1 nvme5n1[1] 64512 blocks super 1.2 [2/1] [_U]And this is reported to cause a problem.
With the proposed patch, the following messages appear:
[ 227.516807] md/raid1:md0: Disk failure on nvme3n1, disabling device. [ 227.516807] md/raid1:md0: Operation continuing on 1 devices.And mdstat shows only the viable members:
Quoted 9 lines
Quoted 4 lines
# cat /proc/mdstat Personalities : [raid1] md0 : active (auto-read-only) raid1 nvme5n1[1] 64512 blocks super 1.2 [2/1] [_U]I think there's a copy-paste error in this change log (probably my mistake). The first mdstat should have shown nvme3n1 was still there, right?
Possibly. Will be resending.
Cheers,
Hannes
Signature · 9 lines
Dr. Hannes Reinecke Kernel Storage Architect [email protected] +49 911 74053 688 SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer _______________________________________________ Linux-nvme mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-nvme